I agree with you in that. They look functional and utilitarian. The icons have dark, bold colors. The 3D depth is clear and meaningful. There is no platform-wide 'theme' that makes the icons similar to each other, instead allowing the symbol to speak its intention more loudly, and designers to create their own vocabulary within the app. Instead, most icons are outlined in black, making them clear and easily distinguishable.

If you're wrong, we're in it together. I very strongly dislike the modern trend towards flat, unrecognizable widgets and non-descript icons.

I keep hoping it's just a phase, but the economical advantages this trend is bringing looks like something that's going to make it stay with us for a while.

This was before the UX people took over the web and desktop. I blame the introduction of smartphones, which was when the public at large finally came online. To appeal to this group, a lot of interfaces and websites got redesigned, losing a lot of functionality and efficiency. KDE is thankfully still rather usable.

(Actual) HCI experts didn't join the computer industry when smartphones popped up. They did when GUIs became a serious research subject for the industry, back in the 1980s, and a lot of their expertise and effort went into computer interaction throughout the 1990s and 2000s.

Frankly, I cynically suspect that the flat, semi-functional design became widespread when the revenue model of many computer and phone vendors started overwhelmingly relying on super-revolutionary "apps" being churned out on minimal budgets.

Today, you can get a complete design for the kind of money (and time) that got you a really good icon for an OS X application back in 2005. It's a major step back in terms of interaction model, not only looks, but the consumer industry runs on money, not quality.

The problem is that UX became a buzzword, and now unqualified graphics designers are 'practising' it. It's hard to find an actual UX designer, and as a UX designer it's hard to find an actual UX job (as opposed to a job that labels itself as UX but is just graphic design)...

Indeed. That's why I said it's a step back in terms of interaction model, too, not just graphic design. The applications sure look awful, but the interaction is increasingly bad, too (just look at the hordes of UX "experts" who religiously follow the Cult of the Hamburger Menu).

> This was before the UX people took over the web and desktop. I blame the introduction of smartphones, which was when the public at large finally came online.

I doubt the plausibility of that timeline - the original iPhone was announced in January 2007, KDE 4.0Alpha1 was released in May 2007[1].

KDE 4.0 was a terrible release (missing functionality, lots of regressions). IMO, they should have kept it as a develop branch and only released 4.1 as 4.0.

1. https://www.kde.org/announcements/announce-4.0-alpha1.php

No, they should have kept it a development branch until 4.3 or 4.4. It wasn't until then that it was really usable.

What's really funny and sad is that the GNOME team made the exact same mistake with Gnome3.

With the additional issue of too much Gnome-isms creeping into GTK3. The one benefit Qt has is that it is being maintained by a independent entity from KDE. One that intends for Qt to be its own product, and properly cross platform.

4.0 was meant to be a development branch, various distros pushed it out anyway.

People cared about consistent look and feel back then, everyone was hating on java because it got it wrong. Now everything seems to have it's own style and behavior.

True that. Of the three programs I can see on my desktop right now (Chrome, WebStorm and Outlook,) none of them are even remotely trying to look consistent with the rest of the operating system. On the other hand, Office 2000 looks right at home on Windows 2000. Even though some controls were technically non-native, they still fit in because the 3D bevel was universal.

I secretly enjoy the ridicule I get from other web developers for refusing to use Chrome (except for late stage cross browser testing) because it does not even try to fit in.

> People cared about consistent look and feel back then, everyone was hating on java because it got it wrong

That does not jive with my memories - and I remember WinAmp, MusicMatch Jukebox and Nero Burning ROM vividly.

I remember that too, I never said people were consistent. Get the order of menus wrong and you were the devil incarnate, write your own gui toolkit and everyone fondly remember you twenty years on (in the case of winamp, music match and nero were reviled).

> in the case of winamp, music match and nero were reviled

Heresy! WinAmp was much-loved. This is the first time I'm hearing anything but love for WinAmp - it was the best music player/organizer at the time, and they added Milkdrop as bonus

I think he's actually saying that it got winamp loved, but music match and nero reviled.

You should probably mention that the vendors at least make an attempt at a 'standard' for their platforms. Whether or not developers follow it, is up for debate.

https://developer.apple.com/design/ https://material.google.com/ https://developer.microsoft.com/en-us/windows/design

There was a period where there was major demand for form widgets in the browser to have a "native look and feel" to decrease the learning curb for people coming from native apps (while meaningful at the time, is less of an issue when most things are web-deployed), and then like the next month there were demands for CSS styling of form widgets and browsers were being shat on because they only supported it in a limited fashion, only color changes, and not actual shape and text styling.

Now you can make text on a button bold italic, but who knows what that means when you see it in relation to a button that isn't bold italic or is just bold or just italic (other than that the designer has bad taste). I once worked with a designer who wanted control over the focus-caret styling (the dotted line in Firefox/highlighting in Chrome that indicates where keyboard input is destined, which changing has potentially huge impacts for accessibility). Having a UI/UX that grossly visually distinct negatively impacts usability, discoverability, and accessibility, IMO, for very little gain in branding (unless your branding is purposely meant to be "confusingly different from everyone else").

I feel similar. I still can't say if it's nostalgia, these were the things that equated technology, and newness for a computer / visual fan. At the time I hated the distinction UI / content, and was quickly fascinated by CSS like technology, almost lispish through DHTML. Todays interfaces are either boring, tiring or useless to my eyes today. Even though I applauded every steps. Maybe it's a questoin of age, maybe it is useless. I tend to think that too much energy is directed at aesthetics rather than function, and even when it's aiming at function, it's at the cost of simplicity. I feel very happy in text mode with emacs, or in GUI mode with 90s day UIs (be it Mac, Windows 3 - NT5, or mid Gnome / KDE), special mention to the UX of SGI OS, and some multimedia tools such as Discreet compositing tools, Houdini, Maya, Lightwave, Modo who had a bit of both: cute and blocky.

The images you linked to feel very much BeOS to me.

But maybe its just the color palette...

Not that strange, the linked image (http://www.digibarn.com/collections/screenshots/KDE%202-x/kd...) does indeed show something like a BeOS theme (https://regmedia.co.uk/2007/01/30/beos_1.jpg).

Really miss BeOS. Loved the way the desktop worked.

Yeah. On second thought, i think they _are_ using KDE with a BeOS theme